DESCRIPTION: (Applicant's Abstract) This is a proposed extension of a longitudinal study of more than 2000 first-graders first recruited as an epidemiologic sample in 1985-86, and assessed annually through middle school. For this extension of the study, these youths will be interviewed when they are at a median age of 17, 18, and 19 years. The main objectives of this 5 year project are threefold. First, we will test hypotheses about how early onset of drug use might be associated with an increase in risk for drug problems during late adolescence. Second, we will test alternative hypotheses about childhood conditions and processes that might influence levels of risk, protective, and resiliency factors, including conduct problems, affiliations with deviant peers, and possible changes that occur after initial drug-taking (e.g., in parent monitoring levels). The set of possible effect-modifiers and mediators to be studied include hypothesized resiliency factors such as religiosity and social bonds. Third, we will test alternative hypotheses about how tobacco smoking and other drug use might influence the onset of psychiatric disturbances and account for psychiatric comorbidity over time. This work will have a primary focus on depression, self-harm, and suicide attempts. All three of these objectives will be addressed in a coordinated fashion, drawing upon the new waves of standardized data gathering of this longitudinal epidemiologic sample between 1997-2000. The long-term goal of this project is to identify malleable characteristics associated with increased risk for drug involvement in adolescence and later risk of drug dependence syndromes. These characteristics should warrant further evaluation as the targets of preventive interventions in future studies.